On Mon, Oct 07, 2019 at 11:41:59AM +0300, Yurii Monakov wrote:
> ks_pcie_stop_link function never cleared LTSSM_EN_VAL bit so
> link training was never triggered more than once after startup.
> In configurations where link can be unstable during early boot,
> for example, under low temperature, it will never be established.
>
> Signed-off-by: Yurii Monakov <monakov.y@xxxxxxxxx>
> ---
> drivers/pci/controller/dwc/pci-keystone.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/pci/controller/dwc/pci-keystone.c b/drivers/pci/controller/dwc/pci-keystone.c
> index f19de60ac991..ea8e7ebd8c4f 100644
> --- a/drivers/pci/controller/dwc/pci-keystone.c
> +++ b/drivers/pci/controller/dwc/pci-keystone.c
> @@ -510,7 +510,7 @@ static void ks_pcie_stop_link(struct dw_pcie *pci)
> /* Disable Link training */
> val = ks_pcie_app_readl(ks_pcie, CMD_STATUS);
> val &= ~LTSSM_EN_VAL;
> - ks_pcie_app_writel(ks_pcie, CMD_STATUS, LTSSM_EN_VAL | val);
Oh yeah, that doesn't look right to me. Good spot. Thanks for this!
> + ks_pcie_app_writel(ks_pcie, CMD_STATUS, val);
As far as I can work out, this bug existed from the beginning - can
you please resend with this tag?
Fixes: 0c4ffcfe1fbc ("PCI: keystone: Add TI Keystone PCIe driver")
Can you also update the commit subject to emphasise it's a bug fix,
e.g. PCI: keystone: Fix ... or similar.
Thanks,
Andrew Murray
> }
>
> static int ks_pcie_start_link(struct dw_pcie *pci)
> --
> 2.17.1
>
